When Lineage Logistics, the world’s second-largest refrigerated warehouse network, wanted to take a chunk out of its electricity bill, they didn’t turn to a seasoned warehouse manager or logistics pro — they hired a 27-year-old PhD in applied physics from Harvard. Alex Woolf, now one year into his position as principal data scientist at Lineage, overseas how 111 cold storage warehouses, each the size of three Home Depot stores, holds temperature for customers like Conagra, General Mills and Driscoll’s.
Woolf and his team have used basic science and physics to drastically change the energy cost of one of their California facilities, and Woolf says that they’re on track for that facility to generate revenue from its electricity use soon.
We caught up with Woolf at The Mixing Bowl Hub’s Food IT: Farm to Fork Conference back in June to find out how they did it.
Tell me about how you’ve changed the way your warehouses use power?
As scientists, we use math and science to solve industrial problems. For a space like logistics or even food & agriculture, many inefficiencies are based on fundamental constraints which can be mitigated with technology. Our approach is to identify and remove the constraints that lead to pinch points. We then implement a solution that is unhindered by the previous limitation, enabling us to realize greater efficiency. The process continues again to more difficult constraints and higher value solutions.
Lineage is owned by a private equity firm called Bay Grove Capital. They assembled Lineage Logistics through the acquisition of 22 premier logistics organizations to become one of the world’s largest cold chain logistics companies. The best way to think about our company is to look at the warehouse itself. Take Home Depot, multiply the square footage by three, multiply the height by two, drop the temperature to zero degrees Fahrenheit then put a bunch of talented people bouncing on forklifts up and down the aisles.
AgFunder Co-Investment Fund III is now open for investment. Closing June 15, Spots are limited.
There is a multitude of interesting mathematical challenges within the warehouse. There is power-scheduling and labor scheduling. There is asset utilization, or problems related to “How do I put more in the box, cool food faster and otherwise maximally use my asset?” Finally, there are problems outside the warehouse, which ask the questions, “How do I optimally design this transportation network in order to maximize efficiency across customers’ complex cold chains?” In answering these questions, the goal is to help drive waste out of the global supply chain.
And therefore sustainability?
We’re able to drive sustainability goals without ever using the word. Many of our projects related to energy efficiency can be justified on a purely financial basis.
How have you made those massive cold storage facilities more efficient?
Here’s a great example – Lineage has to use power. There’s nothing that we can do to completely eliminate that need, but maybe we can mitigate it as a constraint.
You can go back to first principles and ask, “What drives our cost?” The simple answer is labor and electricity. Electricity costs are derived from how much energy you use multiplied by the price you pay to use it. The constraint is you must keep the warehouse at or below zero degrees Fahrenheit to keep the food we store safe for consumption.
When you start digging into the data, you will notice the electricity rates are not constant throughout the day. We discovered we were cooling the warehouse like a thermostat, so it is clicking on and off at all hours throughout the day. Our usage was independent of time, but our costs were very time-dependent because when rates go up, we pay more. So we have this idea that we can treat the warehouse like a battery – we can cool the warehouse when rates are cheap, then turn it off and allow the warehouse to slowly heat up to the maximum acceptable temperature without compromising the integrity of the products stored within.
Take, as an example, the expense associated with cooling a 700,000 sq. ft. facility. To test our theory, we would significantly increase the sensor count to more precisely understand the temperature throughout the room. Once we know the temperature throughout the room, we can quantify how much we can cool it and how long it takes to warm up. With that knowledge, we can feed it into an algorithm that also knows today’s electricity rates and computes the optimal cooling schedule. The whole process is automated and lives in the cloud. So now we’re using the cloud to help control warehouse temperature by automatically calculating and pushing the optimum cooling schedule. As a result, we project being able to cut electricity costs up to 30%.
To what extent have you rolled this out?
We are developing this technique in-house and have assembled a deployment roadmap. Our goal is to roll out to 20% of our facilities over the next 18 months.
Right now we spend tens of millions on electricity every year. Our goal is to have a facility where we get paid to use power for the month. We want to cooperate with the grid, and there are times when you can get paid to either use or not use power. So instead of having a facility where we pay to use electricity, we would actually cooperate with the market in such a way that will turn our power usage into a profit center.
Literally. We’re aiming to get a power bill with a plus sign instead of a minus sign on it. Intervals of negative power rates are becoming more and more common due to renewables. During these intervals, the market pays consumers to use electricity because the grid has an oversupply of electricity and needs to burn it off. Conversely, when the sun goes down and solar panels stop generating power, there can be supply shortages on the grid. We could help the grid by reducing our electricity usage during these times. This could save the grid from having to build a peaker power plant, which is only used intermittently throughout the day.
It is understandable to pause and consider the impact of thermally cycling the products within the warehouse. However, we’ve actually quantified the temperatures in the warehouse orders of magnitude better than we ever have before. So, we’re actually ensuring better food better quality, cooperating with the grid and lowering our electricity costs all at once.